PSI - Issue 41

Hendrik Baarssen et al. / Procedia Structural Integrity 41 (2022) 183–191

190

8

Baarsen et al. / Structural Integrity Procedia 00 (2022) 000–000 Table 2: Results of specimen subjected to fatigue loads

Nu , exp Nu

Specimen

A net

LoadAmplitude

R

Cycles

N u

N u , exp

Type of crack

Depth of crack

[mm 2 ]

[kN]

[-]

[kN]

[kN]

[-]

B5 B6 B7 B8 B9

739.68 737.12 734.40 734.80 735.24 738.20 738.24 730.35

191.97 192.15 191.86 193.02 190.84 191,17 192.35 192.19

0.051 159578 287.19 268.88 0.94

Corner Crack (*) Fatigue failure

7.0 mm

0.050 94188 286.19

-

-

0.051 114381 285.14 276.11 0.97 Corner and trough crack 2.0 - 3.0 mm

0.048 100178 285.29

-

-

Fatigue failure

0.054 62394 285.46 334.99 1.17 0.053 48390 286.61 285.02 0.99 0.050 37155 286.63 324.03 1.13 0.050 81968 283.57 275.75 0.97

No crack

B10 B11 B12

Corner crack (**)

1.0 - 3.5 mm

No crack

Corner crack (*)

2.5 mm

(*) on one side (**) on two sides

W

1,50

FAD - Option 1 FAD - Option 2 Experiments

corner crack

c

1,00

B

a

Kr

0,50

hole

through thickness crack

ligament

0,00

0

0,5

1

1,5

Lr

2a

(a)

(b)

Fig. 6: (a) Crack types; (b) Failure assessment diagram

3.3. The prediction of the FAD

The assessment has been carried out considering the crack sizes as measured on the fracture surfaces of the speci mens B5, B7, B10, and B12, and using the failure assessment diagram, as described in Section 2.3. In particular, the crack size and shapes are considered as obtained from Figure 4 and Table 2, taking into account one of the schemes reported in Figure 6a. For the determination of K mat , using Equation 5, the value of the temperature T 27 J is considered as prescribed in part 1-10 of the Eurocode 3 CEN (2005b), i.e. T 27 J = 20 ◦ C. The assessment is repeated considering both Option 1 and Option 2 FAD, and the results are plotted in Figure 6b. The measured average values of f y and f u are used for Option 1 FAD, whereas the average true stress - true strain curve has been used for Option 2 FAD. On average, the ratio between the predicted failure load by the FAD and the experimental failure load is 0.93 for option 1 and 0.95 for option 2, indicating that the prediction by the FAD is slightly conservative, and reasonably accurate.

4. Conclusions

In this paper an experimental investigation of the net cross-section failure of damaged notched plate specimens is presented. Based on the presented results the following conclusions are made: • The pre-cracked specimens exhibit a reduced strength and ductility, as compared to non-pre-cracked specimens. Moreover, their ultimate resistance is slightly lower than calculated with the design rule of Eurocode 3, when the partial factor is omitted.